The Analog Mixed Signal and Photonic IC (AMPIC) Laboratory is directed by Dr. Vishal Saxena. The Lab is equipped with state of the art workstations and instruments for Analog, Mixed-Signal and Photonic Integrated Circuit design and testing. The AMS lab also houses the Linux server cluster supporting the Cadence and Synopsys tools for research and teaching activities.

Current Research Focus

Presently integrated circuits (IC) and systems research and Semiconductor industry are at an inflection point where it faces three key challenges - (1) sustainability of the decades long trend of performance gain through Moore’s scaling, (2) lack of scalability of current computing and data communication architectures to tackle Artificial Intelligence (AI) and Big Data applications while consuming lower energy, and (3) traditional mixed-signal and RF IC architectures have largely matured and need infusion of new ideas for growth. These challenges present opportunities to IC designers to explore integration of beyond-CMOS technologies such as silicon photonics, emerging devices, and exploitation of machine learning and neuromorphic approaches to realize system-level innovation and improved energy-efficiency. My current research pursues these challenges through the following efforts:
  • Hybrid CMOS Photonic integrated circuits and modeling to Harness Light to Transform IC Design
    • CMOS Photonic ICs for energy-efficient Terabit/s rate optical interconnects
    • Hybrid Radio-frequency and Millimeter-Wave Photonic ICs
    • Emerging applications such as Quantum Computing and Photonics for Machine Learning
  • Neuromorphic Computing and Circuits for Edge-AI
    • Neuromorphic Computing Circuits using emerging memory devices; in-memory computing
    • Spiking Neural Network Algorithms for enabling low-power AI at the Edge
  • High-Performance Analog ICs:
    • High-speed data converters, Delta-sigma modulators, Sensor front-ends
    • Serial links: Phase-locked loops and clock-data recovery

Ongoing and Completed Projects

  • Neuromorphic ICs for pattern recognition aplications with integrated chalcogenide CBRAMs (aka memristors)
  • Energy-efficient CMOS receivers and transmitters for 10 Gbps silicon photonic interconnects
  • CT ΔΣ ADCs with >1 GHz sampling rate in 130nm CMOS
  • Hybrid CT ΔΣ and Multi-step ADCs for wideband (>40 MHz) conversion
  • CT ΔΣ modulator for biomedical applications in DARPA/MITLL 150nm FD-SOI CMOS

Sponsors


Acknowledgements